Syllogimous v4 update

[Fredo Corleone]

I recently took some time to rewrite Syllogimous using TypeScript and Angular, aiming to make it more accessible for contributions and easier to extend. Feel free to try it and take a look at its code. Enjoy!

Try Syllogimous v4

GH Repository

[Bruno 448]

awesome work

[Luis Anaya]

Fredo this is awesome. Super entertaining!

[Jet]

Absolutely fantastic and a step up from v3, but I've encountered a few problems; you might want to look into your syllogisms, Fredo. Plus, the timer is undefined most of the time when the adaptive setting is used. Otherwise, top-notch quality!

[Fredo Corleone]

I know about the undefined timer and I'm trying to replicate it, soon I'll fix it!

[Jet]

Hi Fredo, I believe you'll find that the issue lies within how you're calcuating the stats of the last ten questions for some question type. As of now, only ten questions are evaluated per type as a whole when it should be ten questions for each number of premises within each type. Relying on the "last10Count" property instead of the current question index to determine whether the ten-question threshold has been reached would amend this behaviour. Hope you find this useful!

[Fredo Corleone]

Indeed you got it right! I've found some time to fix it this morning! Try again :)

[Fredo Corleone]

Playground mode is now available! You can configure the settings to play just as you did in v3, adjusting them however you like.

[Ledroit]

thank you fredo !! 

[Aldona Odojewska]

Add Stroop mode as in v3.

--
You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brain-trainin...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/brain-training/8876653e-bfa5-47f1-b2f3-3f9dd890c591n%40googlegroups.com.

[Jet]

Fredo, I'm 99% sure that there's been a fault somewhere in the implementation of meta relations. I've realised I haven't seen any in the last hundred or so questions.

[Fredo Corleone]

Playground mode right?

If so then you found a bug xD

[Jet]

Unfortunately, the problem persists regardless of mode. There's also another issue in the stats menu as it's showing an incorrect amount of questions completed and rate of accuracy for each question type. Beyond these two, I don't believe I've encountered any other issues - would be awesome to see them fixed.

[Fredo Corleone]

I'll fix these bugs during the upcoming weekend.

[Rich Jackson]

Guys, you do really think that this Syllogimous does have more benefits than Dual/Quad N back?

Any motivation pls? I need to brain training.

čt 19. 12. 2024 v 0:15 odesílatel Fredo Corleone <stopchemt...@gmail.com> napsal:

To view this discussion visit https://groups.google.com/d/msgid/brain-training/7f0bdf75-9b10-4926-b0de-c0158336c4afn%40googlegroups.com.

[Arnav Latkar]

It has more benefits but I think you should do both consistently, everyday, one of them in the morning and one before bed. 

[Rich Jackson]

Does this game   Syllogimous  have infinite questions? I am wondering how it was coded like some word randomizer right? So technically there is infinite questions. 

I am asking if I can keep training like months in row and there will still be diff. question. I guess some randomizer like shuffling cards script.

Dne pátek 20. prosince 2024 v 18:09:42 UTC+1 uživatel Arnav Latkar napsal:

[Fredo Corleone]

Yes, it's randomized so you won't find the same questions over and over.

I've just added two new question types, inspired from the LSAT: Linear Arrangement and Circular Arrangement.
Let me know if you find bugs, as usual thank you!

[Rich Jackson]

Great! 

Dne neděle 22. prosince 2024 v 17:45:39 UTC+1 uživatel Fredo Corleone napsal:

[Arnav Latkar]

Hey Fredo, I didn't find circular and linear arrangement in Syllogimousv4. 

[Fredo Corleone]

Go in playground mode and select only those. In order to see those in the arcade mode you need to level up.

[Ledroit]

hello, how can i put the sound of 3D Dual n back to 3D Hyper n back.
His voice is smooth, it help me focus 

[Fredo Corleone]

It should be easy to preload all the sounds you want and play them instead of the synthetic voice.

DL the code, put the mp3s into the project folder, on top of the file preload the sound files with howler (I think I've used that library) then look for the instances where the voice speaks (look for "utterance") substitute those with a corresponding sound.play()

[Ledroit]

Thank u, i wish you all the best in your life mate 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/a86f87b6-7bdb-4f59-947f-2ce5962087a4n%40googlegroups.com.

[Elina Joy]

syllogmous stopped working ;(

[Fredo Corleone]

It's up and running! I got notified by G that the software was down, so I had to re-deploy it.

I've updated how the software adapts to user skill levels. Previously, it was driven by a matrix of premise tiers and question types. Now, the number of premises is based on accuracy for each question type, while tiers are determined by score. The score simply enables or disables question types.

This new system should be more responsive and adaptive. Have fun!

[Elina Joy]

it's up again! hoorah!

[Jet]

I dig these new changes, Fredo. The latest progression system feels like the most natural approach to scale difficulty. One issue I've noticed, however, is that meta relations are applying negations incorrectly. You may see the problem highlighted below; a simple tweak to the code should fix this behaviour:

[Fredo Corleone]

Thanks for your help. Are you willing to make a PR that fixes the issue? :)

[Alberto Flaño Lombardo]

I tried today again and I just got 10 trues in a row. That is a 1/2^10 probability.

I reloaded the game and it seems to run fine now. Maybe it was indeed random. 

[Fredo Corleone]

My friend, have you ever thought about betting on the roulette at Monte Carlo? xD

[Fredo Corleone]

Jet I should have fixed the issue of negations in meta, I've also changed the wording could you check? Thanks
Ah, a user that goes by the name of canyone2015 on github proposed a new algo for generating syllogisms, I've deployed it. Give it a test please!

Il giorno martedì 14 gennaio 2025 alle 09:45:23 UTC+1 Jet ha scritto:

[Alberto Flaño Lombardo]

Hahaha. In fact, I first thought about how a streak like this one would destroy you if you try to pull off a Martingale.

https://en.wikipedia.org/wiki/Martingale_(probability_theory)

Then I thought about visualizing probabilities in higher order dimensions. 

[Taurean]

Does anyone else have a very hard time with carousel mode. It's essentially rft with a wmc component.

To view this discussion visit https://groups.google.com/d/msgid/brain-training/a7d82562-959a-4f58-973f-9e053f6bfa52n%40googlegroups.com.

[Arnav Latkar]

carousel mode was so difficult that ive basically quit syllogimous. just focusing on DnB rn. Ill come back to syllogimous later, or maybe only do LSAT stuff, cuz that alone trains RFT pretty good. 

You received this message because you are subscribed to a topic in the Google Groups "Dual N-Back, Brain Training & Intelligence" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/brain-training/6VeczwxEPQU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to brain-trainin...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/brain-training/CAPdzpFDms4c9LxZiE0EPzphSBng7DOjbKpBuF1jhomR6cmyFFA%40mail.gmail.com.

[Alberto Flaño Lombardo]

With real words, carousel mode is as easy as the normal mode. 

What is the new algorithm, Fredo? 

[Fredo Corleone]

It creates a graph and checks whether the propositions hold true and results in a set of polysyllogisms. 

[Arnav Latkar]

nope. Its way more difficult. Because u need to hold t all in ur head, or keep flipping back and forth. 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/71133001-4807-4894-8f2d-48506d685f7dn%40googlegroups.com.

[Taurean Lee]

Yeah try relating the 8th premise with the first one without looking at them. Good luck

To view this discussion visit https://groups.google.com/d/msgid/brain-training/CAGMsuaxcMEkzM1vkJ7joBjPCM-a6O9GOutA-KTF-63qjdHab3Q%40mail.gmail.com.

[Fredo Corleone]

Well, relating the last and the first is easy because the carousel wraps around xD

[Alberto Flaño Lombardo]

I don't think I will be needing any luck.

This is me playing the previous version: https://youtu.be/h_gdkr5Bu5M?feature=shared

[Arnav Latkar]

My working memory is lacking so Im just doing Dual N-back for now. Maybe I don't need to do carousel mode because my main goal with syllogimous was just improving fluid reasoning, and my memory is already slowly improving from Dual N-back. I feel like the main thing holding me back is snoring during sleep. Im gonna try losing 40 lbs and get down to 12% bodyfat, and I will see if it gets better. 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/624cc579-7bc9-4fa6-8306-363a0345ec86n%40googlegroups.com.

[Taurean]

pervertid...@gmail.com

Ngl this is one of the hardest cognitive flexes I've ever seen. Good shit. Did you build this over time or no? If so how long did it take. 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/624cc579-7bc9-4fa6-8306-363a0345ec86n%40googlegroups.com.

[Alberto Flaño Lombardo]

12% is too low, man. You don't need to go that low, around 14%-15% is enough to be in good shape and it's sustainable in the long term. Trust me, I went that low for a while and it kills you over time. 14% can be maintained without extreme suffering. I'm at 14% today. To hover around 10% you need to stop enjoying food and have a routine of constant very intense exercise. Unless you want to compete in some kind of sport, I think it is not worth it.

I don't know how long it took. With Robert's app I was able to do 99 premises in this manner in two or three weeks, but the premises were much simpler. I have never really been very consistent with my mental training. I sort of built up this ability over time, but I think I have a good memory by default. 

[Arnav Latkar]

12% is not that lean bro. I was 12% a few years back before I started bulking. 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/1accdc86-d839-499c-bc89-07598d1ced97n%40googlegroups.com.

[Alberto Flaño Lombardo]

It is probably something genetic, ten. I was constantly miserable and cold when I had that body fat percentage. 

[Fredo Corleone]

New question type named graph matching.
Check whether two graphs are isomorphs, the problem is presented as an edge list.

[CLJACKSON04]

Nice work Fredo! I’ve been waiting for something graph mapping test related. Despite all the research, indicating that RFT most consistently boosts fluid intelligence, I’ve always felt that the graph mapping test was more effective. If you’re looking for a good next step to amplify the results of the game check out this article. The TLDR is that college students who were trained to acquire a synthetic abilities by reading roughly 50,000 words with various colored letters experienced a 12 point increase in IQ. If you could set up the game so that you could toggle on and off this feature, it seems like it could be beneficial. 

https://pmc.ncbi.nlm.nih.gov/articles/PMC4235407/#:~:text=Compared%20to%20a%20control%20group,an%20average%20of%2012%20points.

To unsubscribe from this group and stop receiving emails from it, send an email to brain-training+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/brain-training/dc38db52-f757-4b73-a021-b7c9e3ed1c9fn%40googlegroups.com.

[Arnav Latkar]

a bit confusing and tricky lol 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/dc38db52-f757-4b73-a021-b7c9e3ed1c9fn%40googlegroups.com.

[Alberto Flaño Lombardo]

I created a rudimentary synesthetic reader a long ago based on that study.

[Arnav Latkar]

which study? 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/c2486436-bd9c-4d9b-8a57-47210662ec9fn%40googlegroups.com.

[Jet]

Thank you for the quick fix, Fredo. It looks like meta relations are working perfectly now. With that sorted, I'd also like to bring up another issue regarding the training unit accuracy calculations. Specifically, the code is using the default training unit length (10) instead of the value the user has defined. For example, with a unit length of 15 and requiring full accuracy to progress, one would only need to answer ~66% of questions correctly for the amount of premises to increase under the current system. The code in question is attached below.

Also, touching on what Jackson mentioned earlier, I've hacked up a program that procedurally generates puzzles similar to those from the graph mapping test. I'll polish and share it here later if anyone is interested in trying it out.

[CLJACKSON04]

@Jet on the Graph Mapping test-yes do please share.

To view this discussion visit https://groups.google.com/d/msgid/brain-training/35235cd9-b951-491c-9aee-e1a2b6cb60e7n%40googlegroups.com.

[Fredo Corleone]

@Jet thanks for your help gonna fix that soon

[Jet]

Happy to help. There's another, quite major, area of interest as well, as shown in the screenshot below. The "premisesUpThreshold" and "premisesDownThreshold" variables should be swapped between the two conditional checks, and the first condition should compare the "correct" percentage (as is done in the second condition) to check if it is below, rather than above, or equal to the premises decrease threshold.

[Fredo Corleone]

Do you have time for a PR? Anyway keep it recede first, procede last.

[Ledroit]

Is there a way to play the older versions of syllogimus v4, the new versions are buggy 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/a8c66e04-6740-4c4c-8a81-a07ca5d77f32n%40googlegroups.com.

[Fredo Corleone]

Yup, you can play with any point of the commit history but I suggest you not. I'd rather suggest you to point out to bugs so that I or others can fix them.

[Ledroit]

Thanks, alright, i found a bug, when i'm having good answers sometimes they downgrade my premisses levels 

[Fredo Corleone]

You must answer a set number of questions before your performance is assessed. So, it does not matter if your last response was correct; if your overall performance is poor, you may still be downgraded.  

[Ledroit]

No, precisely, I have no wrong answers; I’m stringing together correct ones (I reset the game, so I’m starting with easy questions), which I answer in less than 15 seconds. I understand that you’ve implemented the same scoring system as Dual N-Back, but here, I’ve far exceeded 80%, and yet the opposite effect occurs.

To view this discussion visit https://groups.google.com/d/msgid/brain-training/2d84625e-4964-4cc7-ba24-c0fccff4993dn%40googlegroups.com.

[Fredo Corleone]

I tested the mechanism the other day and it looked OK. Is anyone else having the same issue?

[Ledroit]

Hello guys, is it possible that to create a toggle so we can usage the old version of appreciation of score (yk with the levels, (premisses go up with the levels)) 

[Fredo Corleone]

Sure, I'll implement it!

[Ledroit]

Thanks man !! 

--

You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.
To unsubscribe from this group and stop receiving emails from it, send an email to brain-trainin...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/brain-training/4caa72d9-e9ce-4eda-988c-de7375d1faa4n%40googlegroups.com.

[BX6P]

Is anyone practicing thought awareness meditation, when I do this. Intake of stimuli is increased and my observation skills significantly increase. To practice it simply follow the flow of your thoughts, no matter how long, how twisted, and how confusing  or abstract your thoughts get you just maintain your awareness and follow through

--

You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brain-training+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/brain-training/4caa72d9-e9ce-4eda-988c-de7375d1faa4n%40googlegroups.com.

[David Breneisen]

I've done Open Focus meditations which have tapes specifically on thinking and attention, so kind of similar. I think the metacognition skill is really useful to develop to have that second level of awareness above the thinking itself. It's also interesting to try and be aware of what exactly the mind is doing while performing brain training tasks, although it can cause a break in focus on the task itself.

On Saturday, March 8, 2025 at 10:27:02 PM UTC-5 BX6P wrote:

Is anyone practicing thought awareness meditation, when I do this. Intake of stimuli is increased and my observation skills significantly increase. To practice it simply follow the flow of your thoughts, no matter how long, how twisted, and how confusing  or abstract your thoughts get you just maintain your awareness and follow through

On Friday, March 7, 2025, Fredo Corleone <stopchemt...@gmail.com> wrote:

Sure, I'll implement it!

Il giorno mercoledì 5 marzo 2025 alle 23:11:09 UTC+1 Ledroit ha scritto:

Hello guys, is it possible that to create a toggle so we can usage the old version of appreciation of score (yk with the levels, (premisses go up with the levels)) 

--

You received this message because you are subscribed to the Google Groups "Dual N-Back, Brain Training & Intelligence" group.

To unsubscribe from this group and stop receiving emails from it, send an email to brain-trainin...@googlegroups.com.

[Rem]

Hi, Fredo, grat work as allways. I found some bugs:

1. In direction 3d temporal most answers are wrong in cardinal condition.

2. When playing in playground mode, random questions from other modes you haven't selected appear. 

And my suggestion is that the sliders in the options menu are too long. I often accidentally change settings when trying to scroll vertically. I think they should be shorter."

[Rem]

Also why this is wrong. I used many IA they say it sould be correct or when pressed are comming up with some mixed explanations.

[Fredo Corleone]

1. I have to fix some conditions on 4D questions;

2. I think the bug is because it prioritize diametrical opposition over # of steps.

Thanks for your feedback will fix it, including the sliders.

However, the fact that you see questions you haven't picked is because you are using either binary or analogy which are not basic question types themselves and use other question types inside.

[Rem]

Thanks. I dont have analogy or binary selected. I think it has something to do with going to history to see my mistakes and than making back gesture on the screen. But its random.

[Rem]

It looks like it's going into normal mode then and it counts points. But it's random. I have to observe this behavior more to see what triggers it

[Fredo Corleone]

If you can find the causing agent it's top!
Anyway I should have fixed both circular arrangement and 3d temporal can you test and confirm? :D
Thanks

[Rem]

My bad, it's all bueno so far. Good job.

[Fredo Corleone]

Import/export functions now available.

[doe]

Wasn't sure how else to reach out, but I've been using your 3d dual (https://4skinskywalker.github.io/3D-Dual-N-back-Cubeception/) for a bit now and it seems to be the perfect way to increase load on spatial WM. I was looking to do quad with it, but there's no option. I think 3d quad would be even better than regular quad by a mile due to the increased spatial load. If you could add quad to it as an option it would be huge. (I've looked at hyper but I don't like it nearly as much as v0)

[Rem]

+1 I realy like this one and I have never seen it before it has potential. Can you just add piano sounds to it for now? I have been thinking about making musical dual-n-back, to train my ear more, position, notes, scales, instruments etc. But I like to play your verision.

[Rem]

By the way, circular arrangements and graph matching are my favorites now while playing syllogimous. Good additions.

[Rem]

I have found 2 other wrong cases in graph mathcing and 3d temporal. Asking AI to solve them, showing some of the wrong cases, shows me that it is not on my level yet in solving syllogisms.:D

 

[Rem]

its look like direction3D temporal is still incorrect in most cases where carditional position is in conclusion. Here are two more wrong cases one right after the other.

[Fredo Corleone]

Thanks for your help much appreciated. I'll try and fix them soon! Hope these are infrequent cases

[Fredo Corleone]

In the top-right corner of the history card you can click to copy the data of the question, could you send me those? Thanks

[Rem]

{
  "isValid": false,
  "premises": [
    "<span class=\"subject\">Cardboard</span> comes from <span class=\"subject\">Straw</span>",
    "<span class=\"subject\">Widow</span> goes to <span class=\"subject\">Turbine</span>",
    "<span class=\"subject\">Cardboard</span> goes to <span class=\"subject\">Widow</span>"
  ],
  "conclusion": [
    "<span class=\"subject\">Postman</span> comes from <span class=\"subject\">Aubergine</span>",
    "<span class=\"subject\">Knight</span> comes from <span class=\"subject\">Bison</span>",
    "<span class=\"subject\">Bison</span> comes from <span class=\"subject\">Postman</span>"
  ],
  "createdAt": 1742829166386,
  "answeredAt": 1742829188625,
  "negations": 0,
  "metaRelations": 0,
  "timerTypeOnAnswer": "0",
  "userScore": 0,
  "playgroundMode": true,
  "rule": "",
  "bucket": [],
  "buckets": [],
  "coords": [],
  "coords3D": [],
  "graphPremises": [
    [
      "Cardboard",
      "←",
      "Straw"
    ],
    [
      "Widow",
      "→",
      "Turbine"
    ],
    [
      "Cardboard",
      "→",
      "Widow"
    ]
  ],
  "graphConclusion": [
    [
      "Postman",
      "←",
      "Aubergine"
    ],
    [
      "Knight",
      "←",
      "Bison"
    ],
    [
      "Bison",
      "←",
      "Postman"
    ]
  ],
  "type": "Graph Matching",
  "instructions": [
    "Check isomorphism between premise and conclusion graphs."
  ],
  "userAnswer": true
}
{
  "isValid": true,
  "premises": [
    "<span class=\"subject\">Jug</span> is at the same time, one step South and one step East of <span class=\"subject\">Wedding</span>",
    "<span class=\"subject\">Canvas</span> is one hour before and in the same cardinal position of <span class=\"subject\">Jug</span>"
  ],
  "conclusion": "<span class=\"subject\">Wedding</span> is one hour after and one step West of <span class=\"subject\">Canvas</span>",
  "createdAt": 1742829025227,
  "answeredAt": 1742829053264,
  "negations": 0,
  "metaRelations": 0,
  "timerTypeOnAnswer": "0",
  "userScore": 0,
  "playgroundMode": true,
  "rule": "",
  "bucket": [],
  "buckets": [],
  "coords": [],
  "coords3D": [
    [
      "Jug",
      1,
      0,
      1
    ],
    [
      "Wedding",
      0,
      1,
      1
    ],
    [
      "Canvas",
      1,
      0,
      0
    ]
  ],
  "graphPremises": [],
  "graphConclusion": [],
  "type": "Direction3D Temporal",
  "userAnswer": false
}

{
  "isValid": true,
  "premises": [
    "<span class=\"subject\">Button</span> is at the same time, one step South and one step East of <span class=\"subject\">Hawk</span>",
    "<span class=\"subject\">Piano</span> is at the same time and one step North of <span class=\"subject\">Button</span>"
  ],
  "conclusion": "<span class=\"subject\">Hawk</span> is at the same time and in the same cardinal position of <span class=\"subject\">Piano</span>",
  "createdAt": 1742834701700,
  "answeredAt": 1742834732726,
  "negations": 0,
  "metaRelations": 0,
  "timerTypeOnAnswer": "0",
  "userScore": 0,
  "playgroundMode": true,
  "rule": "",
  "bucket": [],
  "buckets": [],
  "coords": [],
  "coords3D": [
    [
      "Hawk",
      0,
      1,
      1
    ],
    [
      "Button",
      1,
      0,
      1
    ],
    [
      "Piano",
      1,
      1,
      1
    ]
  ],
  "graphPremises": [],
  "graphConclusion": [],
  "type": "Direction3D Temporal",
  "userAnswer": false
}

{
  "isValid": true,
  "premises": [
    "<span class=\"subject\">Island</span> is one hour after and in the same cardinal position of <span class=\"subject\">Jungle</span>",
    "<span class=\"subject\">Paste</span> is at the same time and one step North of <span class=\"subject\">Island</span>"
  ],
  "conclusion": "<span class=\"subject\">Paste</span> is one hour after and in the same cardinal position of <span class=\"subject\">Jungle</span>",
  "createdAt": 1742834454040,
  "answeredAt": 1742834564309,
  "negations": 0,
  "metaRelations": 0,
  "timerTypeOnAnswer": "0",
  "userScore": 0,
  "playgroundMode": true,
  "rule": "",
  "bucket": [],
  "buckets": [],
  "coords": [],
  "coords3D": [
    [
      "Island",
      0,
      0,
      1
    ],
    [
      "Jungle",
      0,
      0,
      0
    ],
    [
      "Paste",
      0,
      1,
      1
    ]
  ],
  "graphPremises": [],
  "graphConclusion": [],
  "type": "Direction3D Temporal",
  "userAnswer": false
}

[Rem]

{
  "isValid": true,
  "premises": [
    "<span class=\"subject\">Door</span> is at the same time and one step East of <span class=\"subject\">Hedge</span>",
    "<span class=\"subject\">Dates</span> is at the same time and one step South of <span class=\"subject\">Door</span>"
  ],
  "conclusion": "<span class=\"subject\">Dates</span> is at the same time and one step East of <span class=\"subject\">Hedge</span>",
  "createdAt": 1742924901730,
  "answeredAt": 1742924923768,

  "negations": 0,
  "metaRelations": 0,
  "timerTypeOnAnswer": "0",
  "userScore": 0,
  "playgroundMode": true,
  "rule": "",
  "bucket": [],
  "buckets": [],
  "coords": [],
  "coords3D": [

    [
      "Door",
      1,
      1,
      1
    ],
    [
      "Hedge",
      0,
      1,
      1
    ],
    [
      "Dates",
      1,
      0,
      1
    ]
  ],

  "graphPremises": [],
  "graphConclusion": [],
  "type": "Direction3D Temporal",
  "userAnswer": false
}

[Fredo Corleone]

I give it a look at you questions, two of them are errors on your part: 

"premises": [
  "<span class=\"subject\">Door</span> is at the same time and one step East of <span class=\"subject\">Hedge</span>",
  "<span class=\"subject\">Dates</span> is at the same time and one step South of <span class=\"subject\">Door</span>"
],
"conclusion": "<span class=\"subject\">Dates</span> is at the same time and one step East of <span class=\"subject\">Hedge</span>",

dates is one step east of hedge

"premises": [
  "<span class=\"subject\">Jug</span> is at the same time, one step South and one step East of <span class=\"subject\">Wedding</span>",
  "<span class=\"subject\">Canvas</span> is one hour before and in the same cardinal position of <span class=\"subject\">Jug</span>"
],
"conclusion": "<span class=\"subject\">Wedding</span> is one hour after and one step West of <span class=\"subject\">Canvas</span>",

wedding is one hour after and one step west of canvas

As for the "same cardinal position" issue, I generated 10K random Direction 3D Temporal questions and I wasn't able to produce a single question in which there was "same cardinal" in the conclusion and the 3D coords matrix didn't match, I used the following code:

for (const q of qs) {
    if (hasSameCardinal(q.conclusion)) {
        const [a, b] = extractSubjects(q.conclusion);
        const [ax, ay, az] = getCoords(q, a);
        const [bx, by, bz] = getCoords(q, b);
        if (ax !== bx || ay !== by) {
            console.error(q);
        }
        if (ax === bx && ay === by) {
            console.warn("CORRECT");
        }
    }
}

I'm now wondering what might be the issue, perhaps your browser cached the previous version of the scripts (this issue was fixed 2 weeks ago).

I'm currently looking into the isomorphic graph issue.

[Rem]

I think I am correct. They only have partial correct positions.

dates is one step east AND one step south of hedge
wedding is one hour after and one step west AND one step north of canvas

Becouse you have conclusions which have 100% exact correct position  like: Conclusion

Beret is one hour after, one step South and one step East of Chalk

[Fredo Corleone]

Cardinality in logic is such that even if you say something is east when it is actually northeast, it is still considered true.

[Rem]

Ohh, so its different than in 3d direction. Got it now. So its acctualy easier and faster to solve it than I thougt. AI did me dirty again. Thank you.

[Wilma Skillma]

People keep finding new fancy ways of training dual n back to procrastinate on actually training n-back for 1+ hour a day cause it's boring. No those fancy training methods don't work any better. Just admit to yourself you don't want to train the necessary amount consistently to.improve.

To view this discussion visit https://groups.google.com/d/msgid/brain-training/e13de8b5-cd5e-4c92-bafa-fcd781624acen%40googlegroups.com.

[Fredo Corleone]

”Ohh, so its different than in 3d direction.” it's the same function with a different parameter, I would be surprised if it works any different.

[Rem]

So, I've been playing with 3D direction in Syllogismos v3, and I could swear it required exact positioning. But maybe I've never encountered questions with partial direction in the conclusion there?

Wilma Skillma - nice troll name. I bet you are not on my lvl in qad-n-back

[Fredo Corleone]

"So, I've been playing with 3D direction in Syllogismos v3, and I could swear it required exact positioning..." I've hacked a quick test to check this and you are incorrect: there are 3D Spatial questions with non-exact direction in the conclusion. Follows the code in case you wanna test yourself:
for (const q of Array(1e3).fill(0).map(() => syllogimous.createDirection3D(2, "Direction3D Spatial"))) {
    if (/South/.test(q.conclusion) && !/East/.test(q.conclusion) && !/West/.test(q.conclusion)) {
        console.warn("FOUND ONLY SOUTH", q);
    }
}

Regarding the graph matching issue you found: around 2% of the graph matching questions were previously incorrect. I've observed that the error rate decreases as the number of premises increases.
I've pushed a fix — after testing 1000 questions for each number of premises from 2 to 10, the error rate is now 0%. Could you give it another try? Thanks

[repo...@gmail.com]

Good job! I'm good at finding a needle in a haystack and seeing problems where everyone else doesn't care. :)  I'll be playing it a lot and if I find anything suspicious I'll let you know. But with the new knowledge and those fixes I think everything should be okay.

[BX6P]

Anyone using brain stimulating devices like TCDS or TACS? Studies show memory, learning and mood are all improved.

To unsubscribe from this group and stop receiving emails from it, send an email to brain-training+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/brain-training/ba288705-601b-4e12-b7df-30cd5dc15335n%40googlegroups.com.

[fruitsnackattack]

what a load of hogwash, actually play the game consistently before you make judgements like that

[Wilma Skillma]

That's what I am saying bro. Train consistently before trying fancy stuff. 

To view this discussion visit https://groups.google.com/d/msgid/brain-training/62317087-6061-4cdc-9d0d-c2643f278e42n%40googlegroups.com.

[Rem]

HI, after playing for some time in playground mode with different modules, I think there is an issue with the number of premises used for graph matching. In playground mode when I select 3 premises, the question sometimes has 2, 3, or 4 premises, it's random. I've also checked with 4 premises and the same problem occurs, the question is presented with 3, 4, or 5 premises.

[fruitsnackattack]

according to chat gpt and all the information i put into it about syllogimous v3 and v4 it says these games can in fact boost IQ

[David Breneisen]

Using chat gpt to answer questions like that for you can lower IQ though.